[0001] Field of Invention. The present invention relates to the methods using signal molecules
to allosterically regulate activity of proteins possessing CRISPR-Associated Rossman
Fold (CARF) domain.
[0002] The invention is set out in the appended set of claims. A method using cyclic or
linear adenylates for activation and/or regulation of proteins possessing CRISPR-Associated
Rossmann Fold (CARF) domain is provided. Exemplary enzymes having the CARF domain
include StCsm6 and StCsm6' from
Streptococcus thermophilus, and TtCsm6 from
Thermus thermophilus.
[0003] In one embodiment, the method uses a novel catalytic activity of the Cas10 protein
subunit in the target RNA-bound Csm complex of the Type III CRISPR-Cas system to convert
adenosine triphosphate (ATP), or linear oligoadenylate triphosphate precursors, to
cyclic oligoadenylates which can be used for allosteric activation of CARF family
Csm6 ribonuclease. The reaction is shown schematically both below and with additional
detail in FIG. 11:

[0004] The conversion occurs with >95% efficiency, and in a short reaction time. Cyclic
oligoadenylates ranging from 2 to 6 are obtained. Enzymes possessing the Palm domain
with a GGDD (SEQ ID NO: 1) motif, such as Csm complexes from other organisms or related
Cmr complexes (16), can be used to produce cyclic oligoadenylates. Cas10 protein contains
an N-terminal HD-domain, two small α-helical domains, and two Palm domains that share
a ferredoxin-like fold with the core domain of nucleic acid polymerases and nucleotide
cyclases. Cas10 contains two active sites: the nuclease active site in HD-domain,
responsible for target RNA-activated ssDNA hydrolysis, and the GGDD (SEQ ID NO: 1)
motif in one of the Palm domains (16-19). The inventors have demonstrated here that
GGDD-active site of Cas10 is responsible for adenylate synthesis from ATP or linear
oligoadenylate triphosphate precursors.
[0005] The method demonstrated, in three Csm6 proteins, that cyclic oligoadenylates act
as allosteric regulators of CARF-family Csm6 ribonuclease. This method is shown schematically
both below and with additional detail in FIG. 26:
inactive CARF-protein + cyclic oligoadenylates (2-6) → active CARF-protein
[0006] In this method other enzymes possessing the CARF domain can be activated by cyclic
oligoadenylates produced by a Cas10-containing complex. The Csm6 CARF-family nuclease
is comprised of two domains: CARF family effector domain and HEPN-family ribonuclease
domain (30). The core of the CARF domain is a six-stranded Rossmann-like fold with
the core strand-5 and strand-6 forming a β-hairpin. The main regions of sequence conservation
are associated with strand-1 and strand-4 of the core domain: the end of strand-1
is often characterized by a polar residue, typically with an alcoholic side chain
(S/T),whereas immediately downstream of strand-4 is a highly conserved basic residue
(K/R) often associated with [DN]X[ST]XXX[RK]signature (SEQ ID NO: 2) (8). CARF domains
occur fused not only to RNases but also DNases, membrane-associated protein domains,
TIM barrel adenosine deaminase Ada domain (8); more domain combinations are likely
to be found. Such proteins could be allosterically regulated by cyclic oligoadenylates
or similar compounds, expanding the inventive method to their regulation.
BRIEF DESCRIPTION OF THE FIGURES
[0007] The patent or application file contains at least one drawing executed in color. Copies
of this patent or patent application publication with color drawing(s) will be provided
by the Office upon request and payment of the necessary fee.
FIGS. 1A, B, C, D, and E show StCsm mediated conversion of ATP to the reaction products.
FIGS. 2A, B, C, D, and E show RNase-activity-deficient (Csm3 D33A mutant) ternary
StCsm complex mediated conversion of ATP to the reaction products.
FIGS. 3A and B show target RNA sequence requirements for the StCsm mediated conversion
of ATP to the reaction products.
FIGS. 4 A, B, C, and D show the effect of mutations in StCsm Cas10 HD and Palm domains
on the conversion of ATP to the reaction products.
FIG. 5 shows metal ion requirements for the StCsm mediated conversion of ATP to the
reaction products.
FIG. 6 shows StCsm binding selectivity for various nucleoside triphosphates (NTPs).
FIG. 7 shows StCsm reactions on different NTPs.
FIGS. 8A, B, and C show StCsm mediated AMP reaction with ATP, 2'dATP and 3'dATP.
FIGS. 9A, B, C, and D show characterization of the StCsm mediated ATP reaction products
by liquid chromatography ESI-MS analysis.
FIGS. 10A and B show characterization of the HPLC-purified oligoadenylates by liquid
chromatography ESI-MS analysis.
FIG. 11 shows a proposed model for StCsm-mediated ATP polymerase and cyclase reaction
mechanism.
FIGS. 12A, B, C, and D show StCsm mediated conversion of synthetic linear oligoadenylates
triphosphate into the corresponding cyclic oligoadenylates.
FIGS. 13A, B, C, and D show the effect of P1 nuclease on StCsm mediated ATP reaction
products.
FIG. 14 shows the effect of nuclease PDE12 hydrolysis on the StCsm mediated ATP reaction
products.
FIGS. 15A and B show 5'-end labeling of the StCsm mediated ATP reaction products.
FIGS. 16A, B, C, D, and E show compounds synthesized from ATP by StCsm show. Cyclic
(c) (3'-5') di-adenylate, tri-adenylate, c-tetra-adenylate, c-penta-adenylate and
c-hexa-adenylate are shown schematically.
FIGS. 17A and B show frames and sequence alignment, respectively, of the CARF domain
of each of 5FSH (SEQ ID NO: 3), SeCsm6 (SEQ ID NO: 4), StCsm6 (SEQ ID NO: 5) and StCsm6'
(SEQ ID NO: 6).
FIGS. 18A and B show frames and sequence alignment, respectively, of the HEPN domain
of 4RGP (SEQ ID NO: 7), SeCsm6 (SEQ ID NO: 8), StCsm6 (SEQ ID NO: 9) and StCsm6' (SEQ
ID NO: 10).
FIGS. 19A, B, C, D, E, and F show activation of StCsm6 ribonuclease by cyclic oligoadenylates.
FIGS. 20A, B, C, D, and E show activation of StCsm6' ribonuclease by cyclic oligoadenylates.
FIGS. 21A, B, and C show the nucleic acid specificity of StCsm6 and StCsm6' ribonucleases.
FIG. 22 shows that cyclic oligoadenylates rescue StCsm6 ribonucleolytic activity in
the presence of inhibitory concentrations of ATP.
FIGS. 23A and B shows that c(AMP)6 but not smaller cyclic oligoadenylates stimulates the StCsm6 and StCsm6' ribonucleases.
FIGS. 24A and B shows that c(AMP)4stimulates the TtCsm6 ribonuclease.
FIGS. 25A, B, C, and D show sequence specificity of StCsm6 and StCsm6'; RNA NS (SEQ
ID NO: 11), RNA NSc (SEQ ID NO: 12), and RNA NSd (SEQ ID NO: 13).
FIG. 26 is a schematic overview of inventive embodiments.
FIGS. 27A and B shows that under certain conditions linear, instead of cyclic oligoadenylates,
can be used for stimulation of Csm6 ribonucleases.
DETAILED DESCRIPTION
[0008] Protein cloning, expression and purification. Wild type (wt) and mutant
Streptococcus thermophilus (St) Csm complexes were obtained as described (2). Briefly,
Escherichia coli BL21 (DE3) was transformed with three plasmids: (i) plasmid pCas/Csm, which contains
a cassette including all the
cas/
csm genes (except
cas1 and
cas2) of the Type III-A CRISPR-Cas system from
S. thermophilus DGCC8004, (ii) plasmid pCRISPR_S3, which contains four identical tandem copies of
the repeat-spacer S3 unit flanked by the leader sequence and the terminal repeat,
(iii) plasmid pCsm2-Tag, which contains a N-terminal-StrepII-tagged variant of
csm2 gene. Next, such cells were grown at 37°C in LB medium supplemented with streptomycin
(25 µg/µl), ampicillin (50 µg/µl), and chloramphenicol (30 µg/µl) and expression of
StCsm complex was induced with 1 mM IPTG. The StCsm complex was isolated by subsequent
Strep-chelating affinity and size exclusion chromatography steps.
[0009] Genomic DNA isolated from
S. thermophilus DGCC8004 strain was also used as the template for PCR amplification of the
cas10 gene (1). The resulting PCR product, containing the
cas10 gene, was cloned into pBAD24_C-HSH expression vector via Ncol and Xhol sites. HSH
tag is a combination of StrepII-tag and two 6xHistidine-tags used to purify Cas10
protein using two different affinity chromatography steps. Full sequencing of cloned
DNA fragments confirmed their identity to the original sequences. For purification
of Cas10 subunit,
E. coli DH10B (ara-) bearing wt pCas10-C-HSH plasmid was grown overnight at 16°C in LB medium
supplemented with ampicillin (50 µg/µl) after the induction of Cas10 expression with
0.2 % arabinose. Cas10 protein was purified using chelating Ni-NTA (nickel-nitrilotriacetic
acid) and Strep-Tactin affinity chromatographies. The elutant, containing Cas10, was
dialyzed against 10 mM Tris-HCl (pH 8.5) buffer containing 300 mM KCI, 1 mM DTT, 0.1
mM EDTA, and 50% (v/v) glycerol, and stored at -20°C.
[0010] csm6 and
csm6' genes were amplified separately by means of PCR, using genomic
S.
thermophilus DGCC8004 DNA as a template, and cloned into pJET1.2 vector. Resulting plasmids were
then amplified in
E. coli cells, purified, and cleaved with Eco31I and Pstl in order to clone these genes into
pBAD24_N-HSH expression vector and generate pStCsm6-N-HSH and pStCsm6'-N-HSH plasmids.
Thermus thermophilus csm6 gene (GenBank accession number TTHB152) was synthesized with a C-HSH Tag (General
Biosystems) and cloned into pBAD24_N-StrepII vector, resulting in pTtCsm6-N-StrepII-C-HSH
plasmid. Full sequencing of cloned DNA fragments confirmed their identity to the original
sequences.
E. coli DH10B (ara
-) was transformed with either pStCsm6-N-Tag, pStCsm6'-N-Tag or pTtCsm6-N-StrepII-C-HSH
plasmid. Next, such cells were grown at 37°C in LB medium supplemented with ampicillin
(50 µg/µl) and expression of Csm6 proteins was induced with 0.2 % arabinose. Subsequent
His- and Strep-chelating affinity chromatography steps were employed to isolate StCsm6,
StCsm6' and TtCsm6. The elutants, containing StCsm6, StCsm6' or TtCsm6, were dialyzed
against 10 mM Tris-HCl (pH 8.0) buffer containing 300 mM KCI, 1 mM DTT, 0.1 mM EDTA,
and 50% (v/v) glycerol, and stored at -20°C.
[0011] Plasmids pOAS1, encoding human OAS1 fused with an N-terminal His-Tag, and pPDE12,
encoding human PDE12 stripped for the mitochondrial targeting peptide and containing
a C-terminal His-Tag, were kindly provided by Dr. P. M. Martensen at Aarhus University.
E. coli BL21 (DE3) cells were transformed with either pOAS1 or pPDE12 plasmid and grown at
37°C in LB medium supplemented with ampicillin (50 µg/µl). After protein expression
was induced with 1 mM IPTG, the growth temperature was reduced to 20°C for OAS1 and
25°C for PDE12. PDE12 was purified using chelating Ni-NTA and subsequent ion exchange
(Mono Q XL, GE Healthcare) chromatographies while chelating Ni-NTA chromatography
was sufficient to isolate OAS1. The elutant, containing OAS1, was dialyzed against
10 mM sodium phosphate (pH 7.4) buffer containing 300 mM NaCl, and 50% (v/v) glycerol
while the PDE12-containing elutant was dialyzed against 20 mM Tris-HCl (pH 7.5) buffer
containing 200 mM KCI, 1 mM DTT, and 50% (v/v) glycerol. Both were stored at -20°C.
[0012] Mutagenesis. StCsm containing Cas10 mutations D16A and D575A&D576A were obtained by the Quick
Change Mutagenesis (QCM) protocol [37] and isolated following the procedures described
for the wt StCsm (see above), as described in [1].
[0013] StCsm6 mutations H24A, T107A, N102A+S105A+T107A, Q129A, and R371A+H376A, as well
as StCsm6' mutation R331A+H336A, were introduced into pStCsm6-N-HSH or pStCsm6'-N-HSH
plasmids by means of Phusion Site-Directed Mutagenesis [38]. StCsm6 and StCsm6' proteins
containing point mutations were isolated using the same protocol as for the wt Csm6
(see above). The sequences of wild-type (wt) Cas10, wt Csm6' and wt Csm6 were deposited
in GenBank (accession number KM222358 for the sequence of CRISPR2-cas locus of DGCC8004).
[0014] Gel filtration. Gel filtration was carried out at room temperature on an ÄKTA FPLC system (GE Healthcare)
using a Superdex 10/300 GL column (GE Healthcare), preequilibrated with 20 mM Tris-HCl
(pH 8.5), 0.5 M NaCl, 7 mM 2-mercatoethanol, 1mM EDTA. StCsm6 and StCsm6' protein
samples at 0.5 mg/ml - 0.6 mg/ml loading concentration were prepared in 100 µl of
the above buffer. Elution from the column was monitored by measuring absorbance at
220 nm. The apparent molecular weights of proteins were evaluated from the elution
volume using a series of standards (Gel filtration Calibration Kit from GE Healthcare).
[0015] Nucleotide binding assay. Nucleotide binding assays were performed by incubating different amounts of StCsm
complexes with 10 nM of
32P-radiolabeled-nucleoside triphosphates in the Binding buffer (40 mM Tris, 20 mM acetic
acid (pH 8.4 at 25°C), 1 mM EDTA, 0.1 mg/ml BSA, 10% (v/v) glycerol). All reactions
were incubated for 15 min at room temperature prior to electrophoresis on native 8%
(w/v) polyacrylamide gel (PAAG). Electrophoresis was carried out at room temperature
for 2 h at 6 V/cm using 40 mM Tris, 20 mM acetic acid (pH 8.4 at 25°C), 0.1 mM EDTA
as the running buffer. Gels were dried and visualized by a FLA-5100 phosphorimager
(Fujifilm).
[0016] StCsm mediated synthesis of cyclic oligoadenylates from ATP. The synthesis reactions of cyclic oligoadenylates by StCsm were initiated by adding
10 mM CoCl
2 into a mix of 200 nM StCsm, 200 nM target RNA, 50 µM ATP and 10 nM α
32P-ATP in the Reaction buffer (33 mM Tris-acetate (pH 7.6 at 37°C), 66 mM K-acetate,
0.1 mg/ml BSA) and carried out at 37°C for 1 h or 1.5 h, unless stated otherwise.
StCsm reactions on different nucleoside triphosphates contained 200 nM of ternary
StCsm, 50 µM of the non-labeled nucleotide and 10 nM of the corresponding α
32P-nucleotide. The reactions were stopped by adding 15 mM EDTA. Reaction products were
separated by TLC on PEI Cellulose F plates (Merck) in 0.5 M phosphate buffer (pH 3.5
at 23°C) or a denaturating 24% (19:1 acrylamide:bis-acrylamide) PAAG and visualized
using autoradiography.
[0017] StCsm mediated synthesis of cyclic oligoadenylates from linear adenylate triphosphate
precursors. 200 nM StCsm was mixed with 200 nM target RNA and 17 µM of triphosphate oligoadenylates
(pp(pA)
3-6) in the Reaction buffer (33 mM Tris-acetate (pH 7.9 at 37°C), 66 mM K-acetate, 0.1
mg/ml BSA) and incubated for 20 min at 37°C before initiating the polymerase/cyclase
reactions by adding 10 mM CoCl
2. The reactions were performed for 1.5 h at 37°C.
[0018] HLPC-MS. StCsm mediated ATP reaction products were analyzed using MS. Electrospray Ionization
mass spectrometry (ESI-MS) was performed in negative mode using an integrated HPLC/ESI-MS
system (1290 Infinity, Agilent Technologies / Q-TOF 6520, Agilent Technologies) equipped
with a Supelco Discovery
®HS C18 column (7.5cm x 2.1mm, 3 µm). Elution was performed with a linear gradient
of solvents A (5 mM ammonium acetate in water, pH 7.0) and B (acetonitrile) at a flow
of 0.3 ml/min at 30 °C as follows: 0-2 min, 0% B; 2-22 min, 20% B; 22-25 min, 50%
B, 25-29 min 100% B. Ionization capillary voltage was set to 5000 V, fragmentor -
to 150V.
[0019] Products of the StCsm mediated AMP reactions with ATP, 2'dATP and 3'dATP were analyzed
using Vanquish Binary U HPLC with DAD coupled to Q Exactive PLUS (Orbitrap) with ESI
ion source. HPLC system was equipped with Hypercarb (50mm×2.1mm, 5µ mparticle size;
Thermo Fisher Scientific) column. The chromatography was performed by elution with
a linear gradient of solvents A (1.0 mM NH
4HCO
3, pH=7.8 water/acetonitrile, 95/5 (v/v)) and B (50 mM NH
4HCO
3, pH=9.5 water/acetonitrile, 40/60 (v/v)) at a flow of 0.4 ml/min at 50 °C as follows:
0 % B to 100 % B in 3 min; 100 % B hold for 4 min, 100 % B to 0 % in 0.5 min followed
by 3.5 min reconditioning. ESI-MS data was acquired in MS Scan mode from 200 to 2000
m/z at 35k resolution in negative ionization mode. ESI parameters: Cappilary voltage
2.5 kV; Sheath Gas 35 (Arb); Aux Gas 10 (Arb); Sweep Gas 0 (Arb); Ion Tranfer Tube
Temp 325°C; Vaporizer Temp 275°C.
[0020] Purification of cyclic oligoadenylates. 200 nM StCsm was mixed with 200 nM target RNA and 50 µM ATP in the Reaction buffer
(33 mM Tris-acetate (pH 7.6 at 37°C), 66 mM K-acetate, 0.1 mg/ml BSA) and incubated
for 20 min at 37°C before initiating the reactions by adding 10 mM CoCl
2. The reactions were performed for 1.5 h at 37°C. Next, samples were purified by HPLC.
HPLC was performed at room temperature on Waters Breeze HPLC system using a Discovery
HS C18 Column (15 cm x 10 mm, 5 µm) (Sigma-Aldrich Supelco) pre-equilibrated with
buffer A (100 mM TEAA (pH 7.0)). Samples were fractionated at the at 1 ml/min flow
rate with a linear gradient of B (60% CH
3CN in buffer A) in A (0%-100% of B over 100 ml). Fractions containing different cyclic
adenylates were pooled and the samples were concentrated on a vacuum concentrator
(Eppendorf) prior to ESI-MS analysis.
[0021] Treatment with P1. 12 µM of cyclic oligoadenylate was incubated with 5 mU P1 nuclease (Sigma) in the
P1 reaction buffer (10 mM Tris-acetate (pH 7.1 at 37°C), 1 mM Zn-acetate) at 37°C
for 1 h. 10 µl of such reaction mix was diluted with water and loaded onto HPLC/ESI-MS
system for analysis.
[0022] Treatment with PDE12. To be used as control, 2',5'-oligoadenylates were synthesized by incubating 17 µg/ml
OAS1 with 2 mM ATP and 5 nM α
32P-ATP in the OAS1 reaction buffer (4 mM Tris-HCl (pH 7.8 at 37°C), 0.2 mM DTT, 0.1
mg/ml BSA), supplemented with 0.2 mg/ml dsRNA and 4 mM Mg-acetate, at 37°C for 3 h.
Every 10 min the reaction mix was supplemented with additional 1 mM ATP and 0.25 nM
α
32P-ATP.
[0023] 12 µM of compound (mix of ATP reaction products, linear oligoadenylate (Metabion),
or OAS1 reaction product) containing 7.5 nM of radioactively labeled compound, was
incubated with 1.3 µg PDE12 in the PDE12 reaction buffer (20 mM HEPES (pH 7.0 at 37°C),
1 mM DTT, 1mM Mg-acetate) at 37°C for 1 h. The nuclease reaction products, along with
control lanes containing identical amount of untreated labeled compound, were analyzed
on 24 % (19:1 acrylamide:bis-acrylamide) denaturing PAAG and visualized by autoradiography.
[0024] 5'-labeling reactions. 50 µM of compound (linear oligoadenylate (Metabion) or HPLC purified cyclic oligoadenylate)
was incubated with 0.5 U T4 Polynucleotide kinase (PNK) and 100 nM γ
33P-ATP in the Reaction buffer A (50 mM Tris-HCl (pH 7.6 at 25°C), 10 mM MgCl
2, 5 mM DTT, 1 mM spermidine) at 37°C for 30 min. In case of direct 5'-labeling of
the StCsm ATP reaction mix, which initially contained 50 µM ATP, the sample was diluted
twice and then incubated with T4 PNK under identical conditions. The 5'-labeling reaction
products were analyzed on 24 % (19:1 acrylamide:bis-acrylamide) denaturing PAAG and
visualized by autoradiography.
[0026] Csm6 nuclease assay. StCsm6, StCsm6', and TtCsm6 nuclease assays were conducted in the Reaction buffer
(33 mM Tris-acetate (pH 7.6 at 37°C), 66 mM K-acetate, 0.1 mg/ml BSA) supplemented
with 1 mM EDTA and containing 5 nM of 5'-radiolabeled and 5 nM of unlabeled RNA NS
(unless stated otherwise; see all RNA substrates used in this study in Table 1) and
500 nM of cyclic oligoadenylate mixture or 0.5-500 nM of the HPLC purified cyclic
oligoadenylate effectors or 5-500 nM linear oligoadenylate with various phosphorylation
levels - linear oligoadenylate 5'-triphosphate (ChemGenes); non-phosphorylated, 3'P
(Metabion) or 5'P containing linear oligoadenylates. 5'P-oligoadenylates were obtained
by incubating 3.(3) µM of non-phosphorylated linear oligoadenylate with 0.5 U T4 polynucleotide
kinase and 1 mM ATP in the Reaction buffer A at 37°C for 40 min.
[0027] Reactions were started by adding 0-10 µM Csm6 and carried out at 37°C. The reaction
products were separated on a denaturing 15% (29:1 acrylamide:bis-acrylamide) PAAG
and visualized by autoradiography. StCsm6 nuclease assays with ATP additionally contained
1 mM of ATP in the reaction mixture.
[0028] Data analysis. The Kyplot 2.0 software [45] was used for calculation of StCsm ATP reaction rate
and StCsm6 and StCsm6' ssRNA cleavage efficiency.
[0029] FIG. 1 shows StCsm mediated conversion of ATP to the reaction products. (FIG. 1A)
ATP reaction of an individual StCas10 subunit. The reactions contained 200 nM StCas10,
50 µM ATP, 10 nM α
32P-ATP (Perkin Elmer), 200 nM target RNA S3/2 and/or 5 nM M13mp18 ssDNA. The assays
were performed at 37°C in the Reaction buffer supplemented with 10 mM CoCl
2. Samples were analyzed by thin-layer chromatography (TLC), followed by phosphorimaging.
Reaction products of the binary StCsm complex (Cas/Csm proteins plus crRNA) and ATP
in the absence or in the presence of single-stranded (ss) DNA are presented in (FIG.
1B) and (FIG. 1C), respectively. Reaction products of the ternary StCsm complex (Cas/Csm
proteins plus crRNA and target RNA) and ATP in the absence or in the presence of ssDNA
are presented in (FIG. 1D) and (FIG. 1E), respectively. The reactions contained 200
nM of the binary or ternary StCsm, 50 µM ATP, 10 nM α
32P-ATP and 10 nM M13mp18 ssDNA. The assays were performed at 37°C in the Reaction buffer
supplemented with 10 mM CoCl
2 and the samples were analyzed by TLC, followed by phosphorimaging. Heating α
32P-ATP at 95°C for 60 min was used to generate radiolabeled ADP and AMP, which were
used as a TLC control (marked as lane M). Rate constants of ATP decay, which are presented
bellow each TLC plate, were calculated by fitting single exponentials to the substrate
depletion data. Cartoons above the graphs depict reaction components: protospacer
is grey in between black lines, the complementary strand (matching spacer in crRNA)
is lighter grey, DNA is a black circle.
[0030] FIG. 2 shows RNase-activity-deficient (Csm3 D33A mutant) ternary StCsm complex mediated
conversion of ATP to the reaction products. (FIG. 2A) Reaction components are shown
schematically in the same manner as in FIG. 1. (FIG. 2B) Electrophoretic mobility
shift assay (EMSA) of ATP binding by wild-type (wt) and RNase-activity-deficient ternary
StCsm complexes. The binding reactions contained 2 nM α
32P-radiolabeled, 8 nM non-labeled ATP and the ternary StCsm at concentrations indicated
by each lane. Samples were analyzed in polyacrylamide gel (PAAG) under non-denaturing
conditions. (FIG. 2C) The reactions contained 200 nM of the ternary wt or mutant StCsm,
50 µM ATP and 10 nM α
32P-ATP. The assays were performed at 37°C in the Reaction buffer supplemented with
10 mM CoCl
2 and the samples were analyzed by TLC, followed by phosphorimaging. Heating α
32P-ATP at 95°C for 60 min was used to generate radiolabeled ADP and AMP, which were
used as a TLC control (marked as lane M). (FIG. 2D) Cartoon depicting the experimental
strategy to monitor the temporal control of ATP conversion to the reaction products
by the ternary StCsm. First, StCsm ternary complex is mixed with ATP and RNase cleavage
is triggered by addition of Ca
2+ cofactor. After a certain delay time, the ATP conversion to the reaction products
is initiated by adding Co
2+. (FIG. 2E) Delay time dependence of the relative ATP conversion to the reaction products
by the wt StCsm and RNA cleavage deficient StCsm (Csm3 D33A) variant. The reactions
contained 200 nM of the ternary wt or mutant StCsm, 50 µM ATP and 10 nM α
32P-ATP. The assays were performed at 37°C in the Reaction buffer supplemented first
with 1 mM CaCl
2 and, after a certain delay time, with 10 mM CoCl
2. The samples were analyzed by TLC, followed by phosphorimaging. Reaction rate with
0 delay time was equated to 100%.
[0031] FIG. 3 shows target RNA sequence requirements for the StCsm mediated conversion of
ATP to the reaction products. (FIG. 3A) Cartoons above gels depict ternary StCsm complexes
that differ in the bound target RNA molecules in the same manner as in FIG. 1. The
agarose gels show degradation patterns of 1 nM of circular M13mp18 ssDNA in the presence
of 5 nM ternary StCsm complexes containing different RNAs FIG. 3A is reproduced from
(1). All RNA substrates used here are bound and cleaved by the StCsm ribonuclease
(2). (FIG. 3B) Reaction products of various ternary StCsm complexes with ATP. The
reactions contained 200 nM of the ternary StCsm with different nucleic acids (NA),
50 µM ATP and 10 nM α
32P-ATP. The assays were performed at 37°C in the Reaction buffer supplemented with
10 mM CoCl
2 and the samples were analyzed by TLC, followed by phosphorimaging. Heating α
32P-ATP at 95°C for 60 min was used to generate radiolabeled ADP and AMP, which were
used as a TLC control (marked as lane M).
[0032] FIG. 4 shows the effect of mutations in StCsm Cas10 HD and Palm domains on the conversion
of ATP to the reaction products. (FIG. 4A) Domain architecture of the
S. thermophilus Cas10 protein. HD-domain denotes HD-type phosphohydrolase/nuclease domain (dark grey
or left); two Palm-domains denote polymerase/cyclase-like Palm domains, one of which
contains GGDD-motif (light grey) (SEQ ID NO: 1); D2 and D4 denote α-helical domains
(grey). Conserved active site residues subjected to alanine mutagenesis are indicated
above the boxes. (FIG. 4B) Reaction components are shown schematically in the same
manner as in FIG. 1. (FIG. 4C) EMSA of ATP binding by wt and Cas10-mutant ternary
StCsm complexes. The binding reactions contained 2 nM α
32P-radiolabeled, 8 nM non-labeled ATP and the ternary StCsm at concentrations indicated
by each lane. Samples were analyzed in PAAG under non-denaturing conditions. (FIG.
4D) Reaction products of the wt and Cas10-mutant ternary StCsm complex and ATP. The
reactions contained 200 nM of the ternary mutant StCsm, 50 µM ATP and 10 nM α
32P-ATP. The assays were performed at 37°C in the Reaction buffer supplemented with
10 mM CoCl
2 and the samples were analyzed by TLC, followed by phosphorimaging. Heating α
32P-ATP at 95°C for 60 min was used to generate radiolabeled ADP and AMP, which were
used as a TLC control (marked as lane M).
[0033] FIG. 5 shows metal ion requirements for the StCsm mediated conversion of ATP to the
reaction products. The reactions contained 200nM of the ternary StCsm, 50 µM ATP and
10 nM α
32P-ATP. The assays were performed at 37°C in the Reaction buffer supplemented with
1 mM EDTA, 10 mM Mg-acetate, 10 mM MnCl
2, 10 mM CoCl
2, 0.1 mM NiCl
2, 0.1 mM ZnSO
4, 1 mM Ca-acetate, 1 mM CuSO
4 or 10 mM FeSO
4 and the samples were analyzed by TLC, followed by phosphorimaging. For control, radiolabeled
ADP was generated by mixing α
32P-ATP, T4 PNK and ssDNA S3/2 in PNK A buffer (ThermoFisher) and incubating at 37°C
for 30 min.
[0034] FIG. 6 shows StCsm binding selectivity for various nucleoside triphosphates (NTPs).
NTP binding by binary and ternary StCsm complexes was monitored using EMSA. The binding
reactions contained 2 nM of α
32P-radiolabeled and 8 nM of the corresponding non-labeled nucleotide (ATP, 2'dATP,
3'dATP, GTP, UTP or CTP (Perkin Elmer)) in addition to the StCsm complex (without
RNA, with the target RNA S3/2 or non-specific RNA NS, as denoted above the gels),
at concentrations indicated by each lane. Samples were analyzed in PAAG under non-denaturing
conditions.
[0035] FIG. 7 shows StCsm reactions on different NTPs. (A) Ternary StCsm complex was incubated
with ATP, 2'dATP, 3'dATP, GTP, CTP or UTP radiolabeled nucleotides. The reactions
contained 200 nM of the ternary StCsm, 50 µM of the non-labeled nucleotide and 10
nM of the corresponding α
32P-nucleotide. The assays were performed at 37°C for 60 min in the Reaction buffer
supplemented with 10 mM CoCl
2 and samples were analyzed by TLC, followed by phosphorimaging. (B) The products of
StCsm-mediated 2'dATP reaction were further analyzed by ion pair reverse-phase (RP)
HPLC-MS. On the left panel, HPLC analysis is presented. On the right panels, the isotopic
patterns of ions (without one, (M-H)-, or two (M-2H)2- protons) of the identified
compounds are presented.
[0036] FIG. 8 shows StCsm mediated AMP reaction with ATP, 2'dATP and 3'dATP. TLC analysis
of StCsm reaction products involving 10 nM of α
32P-radiolabeled ATP (FIG. 8A), 2'-dATP (FIG. 8B) or 3'-dATP (FIG. 8C). 200 nM of ternary
StCsm complex were mixed with the nucleotides (concentrations are indicated below
the TLC plate) and incubated at 37°C for 90 min in the Reaction buffer supplemented
with 10 mM CoCl
2 and samples were analyzed by TLC, followed by phosphorimaging. Heating α
32P-NTP at 95°C for 60 min was used to generate radiolabeled ADP and AMP (or corresponding
deoxynucleotides), which were used as a TLC control (first control lane in FIGS. 8A-C,
marked as lane M). In case of AMP reaction with ATP, 2'dATP or 3'dATP, new bands in
TLC plates, corresponding to unknown products, were identified. These unlabeled products
were then prepared incubating ternary StCsm with 25 µM AMP and 25 µM of corresponding
non-radioactive NTP and products were analyzed by HPLC-MS (FIGS. 8D-F), identifying
the reaction products as corresponding dinucleotides, depicted in (FIGS. 8G-I).
[0037] FIG. 9 shows characterization of the StCsm mediated ATP reaction products by liquid
chromatography ESI-MS analysis. (FIG. 9A) Ion pair reverse-phase (RP) HPLC analysis
of StCsm ATP reaction products is presented. (FIG. 9B) Mass spectra of compounds identified
in StCsm ATP reaction mixture. (FIG. 9C) Isotopic patterns of the identified compounds.
(FIG. 9D) Observed molecular masses and yields of the identified putative compounds.
[0038] FIG. 10 shows characterization of the HPLC-purified oligoadenylates by liquid chromatography
ESI-MS analysis. (FIG. 10A) Ion pair RP HPLC chromatogram of StCsm ATP reaction products
is presented. Fractions for three main peaks were pooled separately and analyzed further
by ESI MS/MS analysis. (FIG. 10B) Ion pair RP HPLC analysis and ESI-MS spectra of
the HPLC-purified oligoadenylates.
[0039] FIG. 11 shows a proposed model for StCsm-mediated ATP polymerase and cyclase reaction
mechanism. Cas10 protein, which is part of the StCsm effector complex, possesses two
ferredoxin-like fold domains P* and P, homologous to the Palm domain of nucleic acid
polymerases and nucleotide cyclases. The one Palm domain appears to be catalytically
inactive, whereas the other has a putative active site featuring the conserved GGDD-motif
(SEQ ID NO: 1). Both of them could be capable of binding ATP molecules. ATP binding
in the P site positions the 3'-OH for the nucleophilic attack on the αP atom of the
ATP molecule bound in the P* site. After the initial reaction, the product could reposition
between the Palm domains so that the 3'-hydroxyl group of pppApA dinucleotide could
attack its own triphosphate moiety, yielding cyclic (c) (AMP)
2 (also known as c-di-AMP). Alternatively, the triphosphate moiety of pppApA dinucleotide
could be bound in only one of the Palm domains and ready for attack by a new ATP molecule.
After a trinucleotide is formed, there are two competing ways for the reaction to
go. Under the conditions described herein, formation of c(AMP)
3 was observed as the predominant reaction. In such case the trinucleotide could reposition
within the two Palm domains for an intramolecular cyclisation reaction, in which the
3'-hydroxyl group of pppApApA trinucleotide would attack the 5'-triphosphate group,
resulting in formation of c(AMP)
3. Alternatively, the trinucleotide could polymerize further with another ATP molecule
before the cyclisation reaction would occur, which would result in c(AMP)
4, c(AMP)
5 or c(AMP)
6.
[0040] FIG. 12 shows StCsm mediated conversion of synthetic linear oligoadenylates triphosphate
into the corresponding cyclic oligoadenylates. 200 nM of ternary StCsm was incubated
with 17 µM of triphosphate oligoadenylate - pp(pA)
3 (FIG. 12A), pp(pA)
4 (FIG. 12B), pp(pA)
5 (FIG. 12C) and pp(pA)
6 (FIG. 12D) (ChemGenes)) - in the Reaction buffer supplemented with 10 mM CoCl
2 for 90 min at 37°C. The samples were purified by HPLC and analyzed by ion pair RP
HPLC and ESI-MS.
[0041] FIG. 13 shows hydrolysis of StCsm mediated ATP reaction products by P1 nuclease.
P1 nuclease from
Penicillium citrinum degrades single-stranded RNA (and less efficiently DNA) to nucleoside 5'-monophosphates.
P1 nuclease shows high phosphomonoesterase activity toward 3'-ribonucleotides, but
2'-ribonucleotides are extremely resistant to P1 (3, 4). HPLC purified compounds (tri-adenylate,
tetra-adenylate, penta-adenylate, synthesized by StCsm from ATP, and hexa-adenylate,
synthesized by StCsm from pp(pA)
5 precursor) were tested under P1 nuclease digestion. Next, the samples were analyzed
by ion pair RP HPLC and ESI-MS. HPLC chromatograms are presented in (FIG. 13A). ESI-MS
results are summarized in the Table (FIG. 13B). MW 347.06 was identified in tri-adenylate,
tetra-adenylate, penta-adenylate and hexa-adenylate P1 nuclease digestion reactions
and corresponds to AMP (theoretical MW 347.06). Molecular mass of adenosine-2',3'-cyclic
phosphate, which could result from digestion of the linear oligoadenylate (schematically
presented in (FIG. 13C)), was not detected. Scheme for cyclic-triadenylate compound
hydrolysis by P1 nuclease, which would result in only adenosine monophosphate (AMP),
is presented in panel (FIG. 13D).
[0042] FIG. 14 shows treatment of the StCsm mediated ATP reaction products with nuclease
PDE12. PDE12 exhibits exonuclease activity on 2',5'- and 3',5'-adenylates (5, 6).
The products of reaction containing 200 nM of ternary StCsm, 50 µM ATP and 10 nM α
32P-ATP were mixed with unlabeled products of analogous reaction (total compound concentration
12 µM) and incubated with PDE12 at 37°C for 1 h in the PDE12 reaction buffer. As a
control, a 5'- γ
33P-labeled tri-adenylate (with 3'-5' phosphodiester bonds) and α
32P-labeled 2',5'-oligoadenylates, produced by OAS1 (7), were subjected to identical
PDE12 treatment.
[0043] FIG. 15 shows 5'-end labeling of the StCsm mediated ATP reaction products. (FIG.
15A) The products from ATP reaction, which contained 200 nM of the ternary StCsm and
50 µM ATP, was labeled in the 5'-labeling reaction using T4 PNK and γ
33P ATP (Perkin Elmer). Only traces of linear pApA were detected. (FIG. 15B) 50 µM of
HPLC purified compounds (tri-adenylate, tetra-adenylate, penta-adenylate, synthesized
by StCsm from ATP, and hexa-adenylate, synthesized by StCsm from pp(pA)
5 precursor) were subjected to identical 5'-labeling reactions. Linear oligoadenylates
(ApA, ApApA, (Ap)
3A, (Ap)
4A, (Ap)
5A, (Metabion)) were 5'-radiolabeled in the same way to serve as both control and marker.
All samples were analyzed in PAAG under denaturating conditions, together with the
products gained from reaction containing 200 nM of the ternary StCsm, 50 µM ATP and
10 nM α
32P-ATP.
[0044] FIG. 16 shows compounds synthesized from ATP by StCsm. Cyclic (c) (3'-5') di-adenylate,
tri-adenylate, c-tetra-adenylate, c-penta-adenylate and c-hexa-adenylate are shown
schematically.
[0045] FIG. 17 shows sequence alignment of the CARF domain of StCsm6 and StCsm6'. (FIG.
17A) Domain architecture of StCsm6. Only CARF domain (light grey or left) was used
in the alignment. (FIG. 17B) Sequence alignment of the CARF domain of
Thermus thermophilus Csm6 (TtCsm6, PDB ID: 5FSH (SEQ ID NO: 3)) with StCsm6 (SEQ ID NO: 5), StCsm6' (SEQ
ID NO: 6) and
Staphylococcus epidermidis Csm6 (SeCsm6; SEQ ID NO: 4). Putative ligand binding site residues targeted for mutagenesis
in StCsm6 are denoted by asterisks below the alignment. 4RGP (SEQ ID NO: 7), SeCsm6
(SEQ ID NO: 8), StCsm6 (SEQ ID NO: 9) and StCsm6' (SEQ ID NO: 10).
[0046] FIG. 18 shows sequence alignment of the HEPN domain of StCsm6 and StCsm6'. (FIG.
18A) Domain architecture of the
S. thermophilus Csm6 protein. Only C-terminal region containing HEPN domains (dark grey) and α-helical
region (denoted as 6H [8], light grey) was used in the alignment. (FIG. 18B) Alignment
of
Streptococcus mutans Csm6 (SmCsm6, PDB ID: 4RGP (SEQ ID NO: 7)) with StCsm6 (SEQ ID NO: 9), StCsm6' (SEQ
ID NO: 10) and SeCsm6 (SEQ ID NO: 8). Residues of the active site motif (RXXXXH) (SEQ
ID NO: 14) targeted for mutagenesis in HEPN domain are denoted by asterisks below
the alignment.
[0047] FIG. 19 shows activation of StCsm6 ribonuclease by cyclic oligoadenylates. (FIG.
19A) Domain arrangement of StCsm6. Conserved residues characteristic of the different
domains and subject to alanine mutagenesis are indicated above the boxes. (FIG. 19B)
SDS-PAGE analysis of the purified StCsm6, M - protein mass marker. (FIG. 19C) Size-exclusion
chromatography. The elution profile of the StCsm6 is presented with the calculated
experimental molecular weight 108.9 kDa, corresponding to protein dimer. The size
markers are ferritin (398 kDa), catalase (226 kDa), aldolase (146 kDa), BSA (62.9
kDa), ovalbumin (47.6 kDa), chymotripsin (19.4 kDa), ribonuclease A (15.6 kDa). (FIG.
19D) Dependence of StCsm6 RNase activity on cyclic oligoadenylates. Nuclease assay
was performed at 37°C for 30 min in the Reaction buffer supplemented with 1 mM EDTA,
0-10 µM StCsm6 (at concentration indicated above each lane), and 10 nM 5'-
33P-labeled RNA NS, in the absence or presence of 0.5 µM of the mixture of cyclic oligoadenylates,
produced by StCsm. Samples were analyzed in denaturing PAAG, followed by phosphorimaging.
(FIG. 19E) RNase activity of StCsm6 CARF mutants. Reaction conditions as in (FIG.
19D). (FIG. 19F) RNase activity of StCsm6 HEPN mutant. Reaction conditions as in (FIG.
19D). In the absence of cyclic oligoadenylates, StCsm6 degrades RNA only at high excess
concentrations (1 µM). Addition of mixture of different cyclic oligoadenylates activates
StCsm6 RNase (which then degrades RNA effectively at 1 nM). The lowest concentration
of StCsm6 required for RNA degradation is indicated by a triangle.
[0048] FIG. 20 shows activation of StCsm6' ribonuclease by cyclic oligoadenylates. (FIG.
20A) Domain arrangement of StCsm6'. (FIG. 20B) SDS-PAGE analysis of the purified StCsm6',
M - protein mass marker. (FIG. 20C) Size-exclusion chromatography. The elution profile
of the StCsm6' is presented with the calculated experimental molecular weight 108.1
kDa, corresponding to protein dimer. The size markers are ferritin (398 kDa), catalase
(226 kDa), aldolase (146 kDa), BSA (62.9 kDa), ovalbumin (47.6 kDa), chymotripsin
(19.4 kDa), ribonuclease A (15.6 kDa). (FIG. 20D) Dependence of StCsm6' RNase activity
on cyclic oligoadenylates. Nuclease assay was performed at 37°C for 30 min in the
Reaction buffer supplemented with 1 mM EDTA, 0-10 µM StCsm6' (at concentration indicated
above each lane), 10 nM 5'-
33P-labeled RNA NS, in the absence or presence of 0.5 µM of cyclic oligoadenylates,
produced by StCsm. (FIG. 20E) RNase activity of StCsm6' HEPN mutant. Reaction conditions
as in (FIG. 20D). In the absence of cyclic oligoadenylates, StCsm6' degrades RNA only
at high excess concentrations (1 µM). Addition of mixture of different cyclic oligoadenylates,
same as for StCsm6, activates StCsm6' RNase (which then degrades RNA effectively at
1nM). The concentration of StCsm, which is required for RNA degradation, is indicated
by a triangle.
[0049] FIG. 21 shows the nucleic acid specificity of StCsm6 and StCsm6' ribonucleases. (FIG.
21A) Analysis of StCsm6 cleavage of dsRNA and corresponding ssRNAs in the presence
of cyclic oligoadenylates. Nuclease assay was performed at 37°C for 30 min in the
Reaction buffer supplemented with 1 mM EDTA, 0-10 µM StCsm6 (at concentration indicated
above each lane), 10 nM 5'-
33P-labeled NA and 0.5 µM of the cyclic oligoadenylates. (FIG. 21B) Analysis of StCsm6'
cleavage of dsRNA and corresponding ssRNAs in the presence of cyclic oligoadenylates.
Reaction conditions as in (FIG. 21A). (FIG. 21C) ssDNA treatment with StCsm6 or StCsm6'
in the presence of cyclic oligoadenylates. Reaction conditions as in (FIG. 21A). The
sequences of NA substrates used in this assay are listed in Table 1. Cartoons above
the gels depict substrates. The concentration of StCsm6, required for effective RNA
degradation, is indicated by a triangle.
[0050] FIG. 22 shows that cyclic oligoadenylates rescue StCsm6 ribonucleolytic activity
in the presence of inhibitory concentrations of ATP. StCsm6 RNase is inhibited by
high (1 mM) ATP concentrations. Even a small amount (0.5 µM) of cyclic oligoadenylates
rescues StCsm6 ribonucleolytic activity in the presence of inhibitory (1 mM) concentration
of ATP. Nuclease assay was performed at 37°C for 30 min in the Reaction buffer supplemented
with 1 mM EDTA, 0-10 µM StCsm6 (at concentration indicated above each lane), 10 nM
5'-
33P-labeled RNA NS and the denoted amount of nucleotides. Samples were analyzed in denaturing
PAAG, followed by phosphorimaging.
[0051] FIG. 23 shows that c(AMP)
6 but not smaller cyclic oligoadenylates stimulates the StCsm6 and StCsm6' ribonucleases.
Analysis of StCsm6 (FIG. 23A) and StCsm6' (FIG. 23B) cleavage of ssRNA NS in the absence
or presence of synthetic c(AMP)
2 (c-di-AMP) (Sigma-Aldrich) or HPLC isolated c(AMP)
3, c(AMP)
4, c(AMP)
5 (synthesized by StCsm from ATP) or c(AMP)
6 (synthesized by StCsm from pp(pA)
5 precursor). Nuclease assay was performed at 37°C for 30 min in the Reaction buffer
supplemented with 1 mM EDTA, 0-10 µM StCsm6 or StCsm6' (at concentration indicated
above each lane), 5 nM 5'-
33P-labeled and 5 nM unlabeled RNA NS and 50 nM of the indicated cyclic oligoadenylate
(unless indicated otherwise above the gel). Samples were analyzed in denaturing PAAG,
followed by phosphorimaging. The concentration of StCsm6 or StCsm6', which is required
for effective RNA degradation, is indicated by triangle.
[0052] FIG. 24 shows that c(AMP)
4 stimulates the TtCsm6 ribonuclease. (FIG. 24A) Analysis of TtCsm6 cleavage of ssRNA
NS in the presence of synthetic c(AMP)
2 (c-di-AMP) (Sigma-Aldrich) or HPLC isolated c(AMP)
3, c(AMP)
4, c(AMP)
5 (synthesized by StCsm from ATP) or c(AMP)
6 (synthesized by StCsm from pp(pA)
5 precursor). Nuclease assay was performed at 37°C for 30 min in the Reaction buffer
supplemented with 1 mM EDTA, 0-10 µM StCsm6 or StCsm6' (at concentration indicated
above each lane), 5 nM 5'-
33P-labeled and 5 nM unlabeled RNA NS and 500 nM of the indicated cyclic oligoadenylate.
Samples were analyzed in denaturing PAAG, followed by phosphorimaging. The concentration
of TtCsm6, which is required for effective degradation of RNA, is indicated by a triangle.
(FIG. 24B) c(AMP)
4 is the activator molecule of TtCsm6 ribonuclease.
[0053] FIG. 25 shows sequence specificity of StCsm6 and StCsm6'. (FIG. 25A) RNA substrates
used in this assay. Mapped StCsm6 cleavage positions are indicated by triangles. (FIG.
25B) WebLogo of StCsm6 (1 µM) ribonuclease cleavage patterns of the oligonucleotides
listed in (FIG. 25A) in absence of cyclic oligoadenylates. (FIG. 25C) c(AMP)
6 is the activator molecule of StCsm6 ribonuclease. (FIG. 25D) Analysis of StCsm6 and
StCsm6' cleavage products of ssRNA NSe (SEQ ID NO: 15), which contains all possible
variations of dinucleotide sequences. Nuclease assay was performed at 37°C for 30
min in the Reaction buffer supplemented with 1 mM EDTA, 0-1 µM StCsm6 or StCsm6' (at
concentration indicated above each lane), 5 nM 5'-
33P-labeled and 5 nM unlabeled RNA NSe and 5 nM of c(AMP)
6 (synthesized by StCsm from pp(pA)
5 precursor), if indicated. The samples were analyzed in denaturing PAAG, with RNA
Decade marker (Ambion) (M, first lane), RNA NSe alkaline hydrolysis marker (H, second
lane) and RNA NSe partial RNase A (ThermoFisher) digest (A, third lane) for reference.
Numbers denote the length of adjacent RNA fragments. (FIG. 25E) Sequence of RNA NSe
with mapped StCsm6 and StCsm6' cleavage positions (indicated by triangles). Both StCsm6
and StCsm6' exhibit similar specificities.
[0054] FIG. 26 shows the mechanism of cyclic oligoadenylate signaling in Type III CRISPR-Cas
system. Binding of the target RNA sequence guided by the crRNA of the Cas10-containing
complex triggers the three activities of the complex: (i) Csm3/Cmr4-mediated cleavage
of the transcript itself, (ii) degradation of the corresponding invading DNA by the
HD domain of Cas10, and (iii) synthesis of cyclic adenylates from ATP, carried out
by the Palm domain of Cas10. The resultant cyclic adenylate is a signaling molecule
that is recognized by the sensory CARF domain in Csm6, which in turn activates the
effector HEPN domain of Csm6. Thus activated Csm6 effectively degrades ssRNA, which
could buy the time necessary to ensure the destruction of invasive genome or eventually
lead to cell death.
[0055] FIG. 27 shows that under certain conditions linear instead cyclic oligoadenylates
could be used for stimulation of Csm6 ribonucleases. (FIG. 27A) Analysis of StCsm6
cleavage of ssRNA NS in the presence of linear hexa-adenylate with various phosphorylation
levels (pp(pA)
6 (ChemGenes); (Ap)
6, (Ap)
5A (Metabion); (pA)
6). Nuclease assays were performed at 37°C for 30 min in the Reaction buffer supplemented
with 1 mM EDTA, 0-1 µM StCsm6 (at concentration indicated above each lane), 5 nM 5'-
33P-labeled and 5 nM unlabeled RNA NS and 500 nM of the indicated oligoadenylate (unless
indicated otherwise above the gel). Samples were analyzed in denaturing PAAG, followed
by phosphorimaging. (FIG. 27B) Analysis of StCsm6 cleavage of ssRNA NS in the presence
of linear hexa-adenylate with various phosphorylation levels (pp(pA)
4 (ChemGenes); (Ap)
4, (Ap)
3A (Metabion); (pA)
4). Nuclease assays were performed as in (FIG. 27A). The concentration of StCsm6 or
TtCsm6, which is required for RNA degradation, is indicated by triangle.
I. SYNTHESIS OF OLIGOADENYLATES
[0056] In the Type III CRISPR-Cas systems, multiple Cas proteins and crRNA assemble into
Csm (Type III-A) or Cmr (Type III-B) silencing complexes that provide interference
against invading nucleic acids (1, 2, 9-13). Csm/Cmr complexes function as RNA-activated
single-stranded (ss) DNases that ensure the destruction of foreign genetic elements
while avoiding the degradation of a host's own DNA (1, 10-11, 14-15). When transcription
of phage DNA is initiated, Csm/Cmr complex, guided by the crRNA, recognizes a complementary
target (called a protospacer) in the nascent phage RNA. The RNA transcript binding
by the Csm/Cmr complex triggers target RNA cleavage by Csm3/Cmr4 subunits and simultaneously
activates the ssDNase activity of the Cas10 subunit for
in cis degradation of ssDNA in the transcription bubble. Type III CRISPR-Cas systems avoid
autoimmunity by checking the complementarity between the crRNA 5'-handle that originates
from the repeat and the 3'-sequence flanking the protospacer in RNA target. Base-pairing
between the crRNA 5'-handle and target RNA represses the Cas10 ssDNase activity thus
protecting the host DNA from degradation. Non-complementarity of the crRNA 5'-handle
to the RNA target in the phage RNA signals of a non-self DNA template and activates
the Cas10 ssDNase for degradation (16). The temporal control of ssDNA degradation
by the StCsm complex is achieved through the target RNA cleavage.
[0057] Cas10 subunit (called Csm1 and Cmr2 in the III-A and III-B systems, respectively)
contains an N-terminal HD-domain, two small α-helical domains, and two Palm domains
that share ferredoxin-like fold with the core domain of nucleic acid polymerases and
nucleotide cyclases (17-19). Cas10 contains two putative active sites: the nuclease
active site in HD-domain and the GGDD-(SEQ ID NO: 1)motif in one of the Palm domains
(16). Whereas by now there is a consensus that the nuclease activity of the HD-domain
is responsible for the ssDNA cleavage
in vitro, the role of the GGDD-(SEQ ID NO: 1) motif of the Cas10 Palm domain in Type III-mediated
DNA silencing has remained uncertain. This motif is essential for DNA interference
in vivo and
in vitro by
Staphylococcus epidermidis Csm (SeCsm) complex (12, 20) but is dispensable for the
in vitro ssDNase activity of Csm complex of
Streptococcus thermophilus (StCsm) (1),
Thermus thermophilus (TtCsm) (14) and Cmr complex of
Pyroccocus furiosus (PfCmr) (10). The crystal structures of
P. furiosus Cas10 (PfCas10) alone and the PfCas10/Cmr3 subcomplex show a single ADP, 3'-AMP or
two ATP molecules coordinated together with divalent metal ions by amino acid residues
of the GGDD-(SEQ ID NO: 1) and P-loop motifs in the Palm domains (18). The conservation
of the complete set of catalytic residues typical of Palm domain polymerases and cyclases
implies that the Palm domain of Cas10 should be enzymatically active but the nature
of this activity has remained unknown. The possible ATP-related enzymatic activity
of
S. thermophilus DGCC8004 Cas10 (StCas10) from a Type III-A CRISPR-Cas system was thus investigated.
[0058] StCas10 alone or in the context of the binary StCsm complex [Cas10
1:Csm2
3:Csm3
5:Csm4
1:Csm5
1:crRNA(40 nt)] shows no ATPase or ATP cyclase activity (FIGS. 1A and 1B). However,
ternary StCsm complex, comprised of binary complex bound to target RNA, converts ATP
into a product that migrates faster than ATP but slower than ADP during thin-layer
chromatography (TLC) (FIG. 1D). M13mp18 ssDNA, which is subject to the HD-domain mediated
hydrolysis (1), has no effect on the conversion of ATP to the reaction product (FIGS.
1C and 1E). Moreover, the D33A mutation in Csm3 protein, which prevents target RNA
cleavage by StCsm complex but not target RNA binding (2), does not impair this ATP
reaction (FIG. 2). Therefore, StCsm-mediated conversion of ATP to the reaction product
requires target RNA binding but not target RNA cleavage. Taken together, these data
indicate that formation of a ternary StCsm complex, which includes target RNA, is
the only prerequisite for the StCsm-mediated ATP reaction. Notably, only target RNAs
S3/2 or S3/14, which are complementary to the crRNA spacer but not to the crRNA 5'-handle,
stimulate ATP conversion to the product (FIG. 3B). Complementarity between the 3'-flanking
sequence and the 5'-handle of the crRNA (target RNA S3/3) inhibited the ATP reaction.
Double mismatches in the target RNA sequence (target RNAs S3/7, S3/9 and S3/15) and
target RNA 3'-truncation (target RNA S3/10) but not 5'-truncation (target RNA S3/14)
impaired the ATP reaction (FIG. 3B). Intriguingly, the requirements for the target
RNA molecule to promote the StCsm activity on ATP are similar to the rules that govern
the ssDNase activity of StCsm (1). Moreover, similar to ssDNA degradation, temporal
control of StCsm-mediated ATP reaction, is achieved through the target RNA cleavage
(FIG. 2).
[0059] The ATP reaction, like the DNase activity of the StCas10 HD-domain (1), can be switched
on or off depending on the non-complementarity/complementarity of the crRNA 5'-handle
to the 3'-flanking sequence of target RNA (FIG. 3). Importantly, the D16A mutation
that compromises ssDNA degradation by the StCas10 HD domain has no effect on the ATP
reaction (FIG. 4) while D575A+D576A mutation in the GGDD motif of the StCas10 Palm
domain abrogates ATP conversion into the reaction product (FIG. 4). Therefore, StCsm-mediated
conversion of ATP to the reaction product is dependent on the GGDD (SEQ ID NO: 1)
of Pol-domain and not on the HD active site of StCas10. The ATP reaction, catalyzed
by the GGDD-(SEQ ID NO: 1) domain of StCas10, is dependent on Mn
2+, Co
2+, or Zn
2+ ions (FIG. 5). Taken together, these data demonstrate that the GGDD (SEQ ID NO: 1)
motif of the StCas10 subunit in StCsm is responsible for metal-dependent ATP conversion
into the reaction product and the reaction is critically dependent on the non-complementarity/complementarity
of the crRNA 5'-handle to the 3'-flanking sequence of target RNA.
[0060] The Palm domain of StCas10 subunit in StCsm tightly binds adenosine-containing nucleotides
ATP and 3'dATP with K
d ranging from 10 to 20 nM in the presence or absence of target RNA but shows no significant
affinity towards UTP, GTP and CTP (FIG. 6). Double D575A+D576A mutation in the GGDD
(SEQ ID NO: 1) domain compromises ATP binding (FIG. 4C). 2'dATP is bound by the binary
StCsm complex with the similar affinity as ATP or 3'dATP; however, target RNA binding
into the ternary StCsm complex visibly decreases 2'dATP binding affinity (FIG. 6).
Not surprisingly, no reaction products are formed when GTP, UTP, or CTP are used instead
of ATP (FIG. 7). Unlike ATP, 3'dATP alone does not produce any reaction products while
2'dATP reacts with significantly lower efficiency (FIG. 7); however, all three of
them cross-react with AMP to form respective dinucleotides (FIG. 8). Notably, AMP
alone yields no products (data not shown). This suggests that StCas10 has at least
two adenosine binding pockets that may have slightly different nucleotide binding
modes as revealed by the crystal structures of PfCas10 (21). In the case of AMP reactions,
one pocket presumably accommodates AMP, while another binds ATP, 2'dATP or 3'dATP,
respectively. In conclusion, these experiments reveal following minimal substrate
requirements for the reactions catalyzed by the StCas10 Palm domain: i) it should
contain Ade; ii) 3'-OH in one nucleotide is required for the reaction to occur; iii)
the other nucleotide should contain a triphosphate moiety.
II. PURIFICATION OF OLIGOADENYLATES
[0061] HPLC and ESI-MS analyses were performed to identify the reaction product obtained
in the ATP reaction catalyzed by the StCsm complex. The major ATP reaction product
(63.6%) showed molecular mass of 987.15 Da (FIG. 9). Such Mw could correspond to either
cyclic or linear (containing 2',3'-cyclic phosphate) adenosine monophosphate (AMP)
trinucleotide (tri-adenylate). The reaction product mix also contained compounds with
Mw 1316.20 (corresponding to either cyclic or linear tetra-adenylate, yield 17.7%),
Mw 1645.25 (corresponding to either cyclic or linear penta-adenylate, 8.0%) as well
as traces of Mw 1974.31 (corresponding to either cyclic or linear hexa-adenylate,
0.5%), MW 1005.16 (pApApA or ApApAp, 1.0%) and Mw 836.04 (pppApA or ppApAp or pApApp
or ApAppp, 4.0%) (FIG. 9). MW 1085.13 (ppApApA, ApApApp or pApApAp), Mw 676.11 (pApA
or ApAp) and Mw 658.11 (corresponding to either cyclic or linear di-adenylate) were
also observed, collectively constituting the remaining 5.2%. It was hypothesized that
pppApA is a putative intermediate product, resulting from initial reaction between
two ATP molecules, while ppApApA and pApApA could originate from pppApApA (same as
pApA from pppApA), a putative product of StCsm-mediated reaction between ATP and pppApA.
[0062] HPLC was used to isolate individual ATP reaction products. HPLC was performed at
room temperature on Waters Breeze HPLC system using a Discovery HS C18 Column (15
cm x 10 mm, 5 µm) (Sigma-Aldrich Supelco), pre-equilibrated with buffer A (100 mM
TEAA (pH 7.0)). Samples were loaded and fractionated at 1 ml/min flow rate with a
linear gradient of B (60% CH
3CN in buffer A) in A (0%-100% of B over 100 ml). Three separate peaks were isolated
(FIG. 10A). Fractions containing different oligoadenylates were pooled and the samples
were concentrated by a vacuum concentrator (Eppendorf). The purified samples were
analyzed on an integrated HPLC/ESI-MS system (1290 Infinity, Agilent Technologies
/ Q-TOF 6520, Agilent Technologies) (FIG. 10B). Peak 1 contained compound of MW 1316.21
Da, corresponding to cyclic or linear tetra-adenylate. Peak 2 contained compound of
MW 987.15, corresponding to cyclic or linear tri-adenylate. Peak 3 contained compound
of MW 1645.25, corresponding to cyclic or linear penta-adenylate, with traces of Mw
658.11 (corresponding to cyclic or linear di-adenylate), 987.15 (corresponding to
cyclic or linear tri-adenylate), 1316.21 (corresponding to cyclic or linear tetra-adenylate),
1974.37 (corresponding to cyclic or linear hexa-adenylate). Thus, separate oligoadenylates
produced by the StCsm complex were able to be purified using HPLC.
III. STRUCTURE OF OLIGOADENYLATES
[0063] The reaction products of the ATP reaction were identified. HPLC-MS analysis revealed
that StCsm converts ATP to the products with Mw 987.15 Da, 1316.20 Da, 1645.25 Da
and 1974.31 Da (FIGS. 9A-9C). There are two possible compounds that have these masses:
linear oligoadenylates with 2',3'-cyclic phosphate or cyclic oligoadenylates, respectively
(FIG. 9D). Cyclic adenylates could be generated by ATP polymerization and cyclization
into the oligoadenylate. FIG. 11 shows a proposed model for catalysis mechanism of
such reaction by the StCas10 Palm domains. Cas10 protein, which is a part of the StCsm
effector complex, possesses two ferredoxin-like fold domains P* and P, homologous
to the Palm domain of nucleic acid polymerases and nucleotide cyclases. While one
of the Palm domains appears to be catalytically inactive and the other has a putative
active site featuring the conserved GGDD-(SEQ ID NO: 1) motif, both of them could
be capable of binding ATP molecules (21). ATP binding in the P site positions the
3'-OH for the nucleophilic attack on the αP atom of the ATP molecule bound in the
P* site. After the initial reaction, the product could reposition between the Palm
domains so that the 3'-hydroxyl group of pppApA dinucleotide could attack its own
triphosphate moiety, yielding cyclic (c) (AMP)
2 (also known as c-di-AMP). Alternatively, the triphosphate moiety of pppApA dinucleotide
could be bound in only one of the Palm domains and ready for attack by the 3'-hydroxyl
group of a new ATP molecule. MW could correspond to ppApApA (1085.13 Da) and pApApA
(1005.16 Da), which could result from traces of ADP and AMP in the reaction mixture
or degradation of triphosphate tri-adenylate, supporting that a tri-adenylate intermediate
(pppApApA) forms during the reaction. After this trinucleotide intermediate is assembled,
there are two competing reaction pathways. Under the experimental conditions, formation
of the c(AMP)
3 was a predominant reaction (FIG. 11 and FIG. 10, Peak 2). In such case the trinucleotide
could reposition within the two Palm domains for an intramolecular cyclisation reaction,
in which the 3'-hydroxyl group of trinucleotide pppApApA would attack the 5'-triphosphate
group, resulting in formation of c(AMP)
3. Alternatively, the trinucleotide could polymerize further with another ATP molecule
before the cyclisation reaction occurs, which would result in c(AMP)
4, c(AMP)
5, or c(AMP)
6 (FIG. 11 and FIG. 10, Peaks 1 and 3).
[0064] To confirm the reaction mechanism, linear precursor oligonucleotide triphosphates
(pppApApA, pp(pA)
4, pp(pA)
5, pp(pA)
6) were synthesized and used as substrates for StCsm catalyzed cyclization reaction.
Corresponding triphosphate oligoadenylates were mixed with the ternary StCsm complex
in the Reaction buffer supplemented with 10 mM CoCl
2 and reaction products were analyzed by HPLC-MS. Upon treatment with wt StCsm, pppApApA
was converted into c(AMP)
3 (FIG. 12A). Similarly, pp(pA)
5 was turned into c(AMP)
4, pp(pA)
5 was converted into c(AMP)
5, and pp(pA)
6 was converted into c(AMP)
6 (FIGS. 12B-12D). Therefore, triphosphate oligoadenylates are intermediates of the
StCsm mediated ATP cyclisation reaction and can be converted to the cyclic-oligoadenylates
by StCsm.
[0065] Further, the purified oligoadenylate compounds were subjected to various biochemical
reactions and analyzed them with HPLC coupled with ESI-MS or in a denaturing PAAG.
First, HPLC purified tri-adenylate, tetra-adenylate, penta-adenylate and hexa-adenylate
were treated with P1 nuclease. P1 nuclease from
Penicillium citrinum degrades ssRNA (and less efficiently ssDNA) to nucleoside 5'-monophosphates. Upon
treatment with P1 nuclease, the oligoadenylates were processed entirely into AMP mononucleotides
(FIGS. 13A, 13B and 13D). MS analysis confirmed that neither adenosine nor adenosine
2',3'-cyclic phosphate, which would result from P1 nuclease mediated hydrolysis of
linear oligoadenylate with 2',3'-cyclic phosphate, are produced in this hydrolysis
(FIG. 13C). This result suggests that the oligoadenylates are cyclic and not linear
molecules containing 2',3'-cyclic phosphate. Treatment of the StCsm mediated ATP reaction
products with nuclease PDE12, which exhibits exonuclease activity on 2',5'- and 3',5'-adenylates
[5, 6], revealed that major product tri-adenylate of the StCsm mediated ATP reaction
is resistant to this exonuclease (FIG. 14). Next, the ATP reaction products could
not be labeled in the usual 5'-labeling reaction (FIG. 15B), supporting their cyclic
structure. Moreover, the migration of the main product of the StCsm catalyzed ATP
reaction (cyclic tri-adenylate) in the polyacrylamide gel differs significantly from
migration of linear pApApA (FIG. 15A).
[0066] P1 nuclease exhibits high phosphomonoesterase activity toward 3',5'-ribonucleotides
but 2',5'- ribonucleotides are extremely resistant to it (3, 4). The ATP reaction
products were efficiently degraded by P1 (FIG. 13); therefore, it is likely that all
three of their phosphodiester bonds are 3'-5'. This is further supported by the fact
that 3'-OH but not 2'-OH is prerequisite for the reaction to occur (FIG. 7 and FIG.
8). StCsm catalyzed reactions of ATP and/or 2'dATP but did not produce any products
when reactions contained only 3'dATP (FIGS. 8A-8C). However, all three above mentioned
nucleotides cross-react with AMP to form respective dinucleotides (FIGS. 8A-8C). This
implies that 3'-OH group in adenosine triphosphate is an attacking group in a nucleophilic
attack forming 3'-5' phosphodiester bond in cyclic oligoadenylates.
[0067] Taken together, these data show that in the presence of target RNA the GGDD (SEQ
ID NO: 1) active site in the Palm domain of Cas10 subunit in the StCsm ternary complex
of a Type III-A CRISPR-Cas system catalyzes synthesis of cyclic oligoadenylates (FIG.
16). Small nucleotide-based compounds often act as signaling molecules in various
organisms: cAMP and cGMP are key messengers in both bacteria and higher organisms;
ppGpp and pppGpp act as alarmones in bacteria; nicotinic acid adenine dinucleotide
phosphate (NAADP), adenosine diphosphoribose (ADPR) and cyclic ADP-ribose (cADPR)
are involved in Ca
2+ regulation in higher eukaryotes; diadenosine polyphosphates (Ap
2-6A) and 2',5'-oligoadenylates participate in certain mammalian immune responses (22).
Therefore, it was hypothesized that the cyclic oligoadenylates, synthesised by StCsm
complex of the Type III-A CRISPR-Cas system, could act as signaling molecules in an
anti-phage defense pathway of bacteria. As in all the other described cases there
are complementary proteins that act as sensors for the nucleotide-based messengers
and generate the required response, there should be a corresponding sensory protein
adjacent to the Type III-A CRISPR-Cas system.
IV. CYCLIC ADENYLATES ACTIVATE CARF-DOMAIN RIBONUCLEASES
[0068] Many CRISPR-Cas systems are associated with genes that appear not to be directly
implicated in spacer acquisition, CRISPR transcript processing or interference against
invading nucleic acids (23-27;
K. S. Makarova et al. The basic building blocks and evolution of CRISPR-CAS systems.
Biochemical Society transactions 41 (2013), p. 1392-1400). For example,
csm6 gene, that belongs to the COG1517 family [24-25], is associated with Type III-A CRISPR-Cas
locus of
S. thermophilus DGCC8004 strain.
S. thermophilus StCsm6' and StCsm6 proteins do not belong to the StCsm effector complex that provides
interference against invading nucleic acids (2). Bioinformatic analysis revealed that
StCsm6 and StCsm6' proteins are related (~35% sequence identity) (16). They both have
a typical Csm6 architecture, which includes the N-terminal CARF (CRISPR-asssociated
Rossman fold) domain (FIG. 17), the middle α-helical "6H" domain and the C-terminal
HEPN domain (nucleotide binding domain, characteristic to both prokaryotes and higher
eukaryotes) (8, 28-29) (FIG. 18). Molecular modeling revealed that N-terminal CARF
domains of both StCsm6 and StCsm6' are most similar to the corresponding domain of
the TtCsm6 (PDB: 5FSH) (29) (FIG. 17). The middle (6H) and the C-terminal (HEPN) domains
were found to be the most similar to the structure of Csm6 from
Streptococcus mutans (PDB: 4RGP) (FIG. 18). Structural models reveal that the CARF domain in both StCsm6
and StCsm6' is highly similar. The largest difference between the two proteins is
observed within the HEPN domain, which is more compact in StCsm6' (FIGS. 18D and 18E).
[0069] Previously, it was shown that activity of the TtCsm6 RNase resides in the composite
active site formed by a pair of HEPN domains, whereas the interface of the CARF domains
features a putative ligand-binding site (29). The putative binding site formed by
CARF domains might be the site for binding the cyclic oligoadenylates, produced by
Cas10 of the Csm complex; to test this, docking experiments were performed using the
dimeric structure of TtCsm6 CARF domains as a receptor and adenosine as the ligand.
The nucleoside over nucleotide was chosen to test the compatibility of shape with
the putative binding site while at the same time avoiding potential complications
due to the electrostatic interaction between the nucleotide phosphate(s) and the positively
charged cleft formed by the CARF domains. Docking revealed that two distinct pockets
within the cleft can bind adenosine (FIG. 17D). Since the TtCsm6 structure is a dimer,
there is another symmetry-related pair of pockets. However, as the asymmetric unit
in the crystal structure corresponds to a dimer, the two pairs of pockets are not
identical. That could be the reason why the adenosine was docked only into the two
and not all four pockets.
[0070] CARF domains of both StCsm6 and StCsm6' lack a couple of α-helices at the N-terminus
compared to the CARF domain of TtCsm6 but otherwise are quite similar to it (FIGS.
17E and 17F). Therefore, the docking results for TtCsm6 were considered to perhaps
be relevant to StCsm6 and StCsm6'. Based on StCsm6 and StCsm6' structural models residues
corresponding to the two distinct pockets at the interface of CARF domains were identified,
and those likely contributing to the nucleobase binding (His24 and Gln129 in StCsm6)
were selected. In addition, three other conserved residues were selected that are
involved in forming the bottom of the cleft (Asn102, Ser105 and Thr107 in StCsm6)
(FIG. 17).
[0071] Domains belonging to the HEPN superfamily often exhibit ribonuclease activity and
are commonly found in prokaryotic toxin-anti-toxin (T-A) and abortive infection (Abi)
defense systems (30). It has been proposed that Csm6 and other CARF family proteins
associated with the Type III CRISPR-Cas systems may degrade RNA transcripts of DNA
invaders complementing the DNA- and RNA-targeting endonuclease activities of the Csm
complex (8). Indeed, in
S. epidermidis Csm6 protein is involved in the degradation of RNA transcripts for the late expressed
viral genes (31). Alternatively, it was suggested that Csm6/Csx1 proteins could contribute
to CRISPR immunity by targeting host (i.e., self) transcripts in order to induce dormancy
or promote programmed cell death of the host (28, 29).
[0072] To understand the role of the Csm6' and Csm6 proteins in
S. thermophilus immunity, StCsm6 (FIG. 19B) and StCsm6' (FIG. 20B) were expressed in heterologous
E. coli host and purified. Both StCsm6 and StCsm6', like homologous SeCsm6, TtCsm6, and PfCsm6
proteins (29, 31), exhibited ssRNA degradation activity in a metal-independent manner
(FIGS. 19D and 20D). Alanine replacement of the conserved putative active site residues
(RXXXXH) (SEQ ID NO: 14) located in the HEPN domain impaired the ssRNase activity
of StCsm6 and StCsm6' (FIGS. 19F and 20E). Notably, the StCsm6 and StCsm6' ssRNase
activity was rather weak and observed only at high (micromolar range) protein concentrations.
StCsm6 (and StCsm6') protein, like most of the other COG1517 family proteins, contains
an N-terminal CARF domain that could act as a ligand sensor. It was hypothesized that
cyclic oligoadenylates, produced by the StCsm complex, could act as a ligand for StCsm6
(StCsm6'). Indeed, the mixture of cyclic oligoadenylates effectively stimulated both
StCsm6 and StCsm6' ssRNase activity, reducing the required protein concentration ~1000-fold
(FIGS. 19D and 20D). Addition of cyclic oligoadenylates did not result in any hydrolysis
dsRNA or ssDNA (FIG. 21), confirming StCsm6 and StCsm6' are ssRNA-specific nucleases.
Mutation of conserved amino acid residues H24, N102, S105, T107 or Q129 in the putative
ligand binding pocket in the CARF domain of StCsm6 significantly impaired ligand-stimulated
ssRNase activity (FIG. 19E).
[0073] StCsm6, StCsm6' and TtCsm6 are homodimers in solution (FIGS. 19C and 20C) [29]. Homology
model of the StCsm6 indicates that the ligand binding site is located at dimer interface
of the CARF domains (FIG. 17). It is likely that cyclic oligoadenylate binding to
the CARF domain induces structural rearrangements in the HEPN domain as well, thus
stimulating the ssRNase activity of the StCsm6. This observation indicates that CARF
domain of the Csm6 proteins acts a sensor for the ligand produced by the StCsm complex
of the Type III-A CRISPR system. Biologically relevant ATP concentrations inhibit
the StCsm6 RNase but its activity can be easily rescued by the cyclic adenylates (FIG.
22).
[0074] Next, it was determined which of the cyclic oligoadenylates, synthesized by the ternary
StCsm complex, was the activator of the StCsm6 and StCsm6' nucleases. For this a synthetic
c(AMP)
2 (c-di-AMP) was used and the compounds that isolated from StCsm ATP or triphosphate
oligoadenylate reaction mixtures by HPLC: c(AMP)
3, c(AMP)
4, c(AMP)
5 and c(AMP)
6. Nuclease assays revealed that of all the cyclic adenylates tested, only c(AMP)
6 stimulated ribonuclease activity of StCsm6 and StCsm6' (FIG. 23). Even 0.5 nM (5
nM in case StCsm6') of c(AMP)
6 was sufficient for efficient StCsm6 ribonuclease activity.
[0075] To elucidate if cyclic oligoadenylates synthesized by Type III StCsm effector complex
are universal activators for Csm6 proteins, considering the TtCsm6 ssRNase activity
was also observed only at high (micromolar range) protein concentrations (29), it
was examined in the presence of cyclic oligoadenylates, synthesized by the StCsm ternary
complex. Only c(AMP)
4 stimulated TtCsm6 ribonuclease activity (FIG. 24). This is in agreement with the
docking and structure modelling experiments that show four possible adenosine binding
sites in the dimeric TtCsm6 CARF domain structure (FIG. 17D). Seemingly, the cyclic
oligoadenylates are signaling molecules in various Type III CRISPR-Cas systems; however,
the exact size of the active compound can vary between organisms.
[0076] To investigate whether StCsm6 and StCsm6' ribonucleases possess sequence specificity,
cleavage of 5'-radiolabeled ssRNA substrates with random sequences (FIG. 25A) was
analyzed. Initial analysis suggested preference for dinucleotide with different variation
of purines (FIG. 25B). To test the sequence specificity further, RNA NSe oligonucleotide
was synthesized containing all possible dinucleotide combinations, interspaced by
a CC sequence. StCsm6 and StCsm6' cleavage of this substrate both in the absence and
the presence of c(AMP)
6 was analyzed. In the absence of nucleotide effector and at high nuclease concentrations,
residual cleavage was most prominently visible at GA dinucleotide (FIGS. 25A, 25B
and 25E), as indicated by the initial analysis. Addition of c(AMP)
6 not only stimulated effective RNA NSe hydrolysis at low enzyme concentrations but
also promoted cleavage at other positions, containing a purine nucleotide, especially
AG (FIGS. 25E and 25F).
[0077] Smaller cyclic dinucleotides (c-di-GMP, c-di-AMP or cGAMP) across the three superkingdoms
of life are often encountered as intracellular messengers that convey specific as
well as global signals (32, 33). They are synthesized in a cell by specific enzymes
in response to different stimuli and are recognized by sensor domains embedded within
different effector proteins. For example, c-di-AMP is generated in response to direct
sensing of endogenous branched DNA, making it a checkpoint regulator (34). In vertebrates,
linear 2'-5' oligoadenylates (2'-5'A), ranging from 2 to 30 oligomers, are produced
in response to sensing of double-stranded viral RNA and stimulate latent ribonucleases
(RNaseL) for the degradation of the invading RNA (35). In prokaryotes, signaling systems
utilizing cyclic di-nucleotides have been considerably well studied; however, systems
centered on other oligonucleotides remained largely unknown. The data herein show
that the cyclic hexa-adenylate c(AMP)
6 synthesized by Cas10 protein in the StCsm complex of the Type III-A CRISPR-Cas system
in response to the target RNA recognition acts as a novel messenger that activates
RNA degradation by Csm6 ribonuclease through binding to the sensor CARF domain (FIG.
26), thereby expanding the role of cyclic oligonucleotide signaling to prokaryotic
defense systems.
V. APPLICATIONS
[0078] The synthesis reaction of cyclic oligoadenylates that can be used in the presently
disclosed methods may be conducted
in vitro by the StCsm complex by adding 10 mM CoCl
2 into a mix of 200 nM StCsm, 200 nM target RNA, 50 µM ATP in the Reaction buffer (33
mM Tris-acetate (pH 7.9 at 37°C), 66 mM K-acetate, 0.1 mg/ml BSA) and incubated at
37°C for 1 h. If needed, reaction conditions could be changed: (i) incubation could
be extended or reduced to a time more or less than one hour, for example, 15 minutes;
or (ii) other bivalent metal ions could be used as cofactors, for instance, Mg
2+, Mn
2+, Ni
2+, Zn
2+or Fe
2+ or (iii) both StCsm and target RNA concentrations could be changed, or (iv) different
substrates could be used at various concentrations, e.g., 17 µM short 5'ppp oligoadenylates
could be used instead of 50 µM ATP; or (v) different reaction buffers with different
pH, ionic strength could be used. These compounds are stable and can be stored at
+4°C or -20°C for months and re-thawed multiple times.
[0079] Ternary StCsm complex, comprised of Cas10, Csm2, Csm3, Csm4, and Csm5 proteins, a
crRNA, and a target RNA transcript which is substantially complementary to a portion
of the crRNA was employed. However, other complexes containing Cas10, such as Csm
complexes from other organisms or related Cmr complexes (16), should be capable of
a very similar if not identical activity. Therefore, Cas10-containing complexes other
than the complex described in the exemplary descriptions may be used in the described
methods.
[0080] Given that the StCsm complex is capable of polymerizing two ATP molecules and the
general mechanism of StCsm-mediated cyclic oligoadenylate synthesis (FIG. 11), it
should be able to cyclize any RNA that has 5'-pppA and A-3'OH. 5'-pppRNAs-3'OH, are
naturally occurring in abundance in prokaryotes, and obtained using
in vitro transcription kits. Moreover, eukaryotic RNA also contains a triphosphate moiety
at its 5' side and a 3'-polyA tail, which makes them potentially good targets for
StCsm-mediated RNA cyclization reaction. Stability of RNA, especially when delivered
to cell cultures, has always been an issue. It has been shown that cyclization of
silencing RNA improves its stability (
Abe et al., Dumbbell-shaped nanocircular RNAs for RNA interference. J Am Chem Soc.
129 (2007) p. 15108-15109). Moreover, under described conditions StCsm complex also mediated synthesis of cyclic
oligodeoxyadenylates, extending its applications to DNA as well.
[0081] Since the Cas10-containing complex was able to synthesize these molecules only in
the presence of target RNA, the synthesis of cyclic oligoadenylates (or any cyclic
nucleic acid synthesized in Cas10-dependent manner) could be used as a signal to report
appearance of target nucleic acids
in vitro. Neither ssDNA nor dsDNA can cause the StCsm complex to synthesize cyclic compounds.
However, known techniques can be used to transcribe a desired DNA and thus produce
target RNA for the Cas1 0-containing complex based on a desired DNA. Thus the described
method can be applied for not only RNA but also DNA detection in various samples.
The appearance of cyclic oligoadenylates could be detected using, for example, mass
spectrometry analysis.
[0082] We discovered that a family of CARF-containing nucleases can be activated by cyclic
oligoadenylates. CARF-containing RNases Csm6 from different organisms are activated
by differently sized cyclic oligoadenylates. Small amounts of cyclic oligoadenylates
are required for efficient activation of CARF-containing RNases: 0.5 nM c(AMP)
6 for StCsm6, 5 nM c(AMP)
6 for StCsm6', 50 nM c(AMP)
4 for TtCsm6. However, we found that much higher amount of linear oligoadenylates can
also activate these proteins (FIG. 27). While the RNA degradation is essentially non-specific,
StCsm6 and StCsm6' show preference for purine sites. Stimulated by cyclic oligoadenylates,
StCsm6 is able to degrade ssRNA at subnanomolar concentrations. In general, signaling
systems have a way to remove the effector molecule. Removal of cyclic oligoadenylates
would inactivate this enzyme. For example, the oligoadenylates are degradable by nuclease
P1 (FIG. 13). Therefore, CARF-containing RNases can be used to remove RNA from various
samples in a controllable fashion. Given the nature and size of the cyclic oligoadenylates,
this could allow initiation of total cell RNA degradation in cells as well as complete
ssRNA hydrolysis
in vitro. Often the CARF domain is found combined with different RNases, such as members of
the RelE (23) and HEPN families (30) suggesting that other RNases could also be activated
for RNA degradation
in vitro by cyclic or linear oligoadenylates similar to Csm6.
[0083] Naturally, the cyclic adenylates, required for activation of CARF-containing proteins,
can be synthesized by the previously described enzymatic synthesis, using a Cas10-containing
complex. CARF-containing protein can be coupled with Cas10-containing complex to create
a unique system for
in vitro platforms and
in vivo tools. For instance, such system could be used for detection of selected nucleic
acids, e.g., viral transcripts, in a manner similar to the SHERLOCK platform (36).
In vivo such system could be used to induce cell dormancy at desired time and/or desired
cell by choosing an appropriate target transcript. This target transcript could appear
at certain stage of cell cycle or under certain other conditions. The system allows
assaying the cellular content upon emergence of the selected target transcript. Alternatively,
the system could be used to distinguish between cells that transcribe the target sequence
(either constitutively or occasionally) from the target-free cells. The system could
be applied to induce cell suicide in response to a foreign or malignant transcript,
e.g., viral or lethal.
[0084] All the necessary components could be delivered to the cells either as DNA, RNA,
or proteins, or any combination of them, by transformation, electroporation, transfection,
etc. Genes encoding the necessary components could also be integrated into the genome
of desired cells. When delivered in DNA form, the expression of system components
could be inducible or constitutive, depending on the need. The target transcript,
which is substantially complementary to a portion of the crRNA, could be (i) delivered
as DNA (for instance, expressed from an expression plasmid), (ii) delivered as RNA,
(iii) emerge after infection (for instance, viral) of the cell culture, or (iv) any
endogenous gene product or non-coding RNA of the cell could be selected as a target.
In some examples, in the case of eukaryotic cells, the expression of these components
would be optimized.
[0085] Other CRISPR-Cas systems possess Csm6 analogues, such as Csx1 (28, 40). CARF domains
occur fused not only to RNases but also DNases, membrane-associated protein domains,
TIM barrel adenosine deaminase Ada domain (8) and more domain combinations might be
found. It is likely that such proteins could be allosterically regulated by cyclic
oligoadenylates or similar compounds; expanding the inventive methods to their regulation.
Therefore, while
S. thermophilus and
T. thermophilus Csm6 proteins are used in the above exemplary descriptions, the invention is not
limited to this source of proteins and other proteins may be used in the described
methods.
[0086] We demonstrated here that cyclic or linear oligoadenylates could be used to activate
ribonucleolytic activity of different Csm6 proteins through stimulation of their CARF-domain.
Often the CARF domain is found combined with different effector domains. Most of the
CARF domain proteins contain a winged HTH (wHTH) DNA-binding domain immediately C-terminal
of CARF, for example in CRISPR-Cas associated protein Csa3 (8, 41). It has been suggested
that such proteins are allosterically controlled transcriptional regulators (41).
Therefore, CARF-containing DNA-binding proteins could be used to regulate gene expression
in a variety of both prokariotyc and eukaryotic cells using cyclic or linear oligoadenylates.
[0087] Many CARF-domain proteins possess additional not only RNase, but also DNase domains,
in particular those of the restriction endonuclease, for example protein VC1899 from
Vibrio cholera (8, 25). Therefore, CARF-containing DNases could be used specifically to cleave target
DNA seguence or non-specifically to remove DNA from various samples in a controllable
fashion
in vitro.
[0088] CARF domains are also found in RtcR proteins, sigma-54 RNA polymerase dependent regulators
of the
rtcBA tRNA and mRNA splicing and repair operon (8, 42, 43). Binding of RtcR protein near
promoter of
rtcBA activates transcription of two proteins, RNA ligase RtcB (44) and 3'-terminal phosphate
RNA cyclase RtcA (45). It was demonstrated that in
Escherichia coli bacteria expression of
rtcAB is activated by agents and genetic lesions which impair the translation apparatus
(42). Therefore, CARF-containing RtcR-like protein could be used to regulate protein
translation in a variety of both prokariotyc and eukaryotic cells using cyclic or
linear oligoadenylates.
[0089] Sometimes CARF domain is fused to a TIM barrel adenosine deaminase Ada domain the
enzyme that catalyzes deamination of adenosine to inosine in the purine salvage pathway
(8). Therefore, CARF-containing deaminase protein could be fused to Cas9 or Cpf1 cleavage
deficient or nicking variants to engineer base editor regulated with cyclic or linear
oligoadenylates via CARF-domain
in vitro.
References
[0090]
- 1. Kazlauskiene et al., Spatiotemporal Control of Type III-A CRISPR-Cas Immunity: Coupling
DNA Degradation with the Target RNA Recognition. Mol Cell, 2016. 62(2): p. 295-306.
- 2. Tamulaitis et al., Programmable RNA shredding by the type III-A CRISPR-Cas system
of Streptococcus thermophilus. Mol Cell, 2014. 56(4): p. 506-17.
- 3. Fujimoto et al. Identity of Phosphodiesterase and Phosphomonoesterase Activities with
Nuclease P1 (a Nuclease from Penicillium citrinum). Agricultural and Biological Chemistry,
1974. 38(4): p. 785-790.
- 4. Fujimoto et al. Specificity of Nuclease P1. Agricultural and Biological Chemistry,
1974. 38(9): p. 1555-1561.
- 5. Kubota et al., Identification of 2'-phosphodiesterase, which plays a role in the 2-5A
system regulated by interferon. J Biol Chem, 2004. 279(36): p. 37832-41.
- 6. Poulsen et al., Human 2'-phosphodiesterase localizes to the mitochondrial matrix with
a putative function in mitochondrial RNA turnover. Nucleic Acids Res, 2011. 39(9):
p. 3754-70.
- 7. Poulsen et al., Enzyme assays for synthesis and degradation of 2-5As and other 2'-5'
oligonucleotides. BMC Biochem, 2015. 16: p. 15.
- 8. Makarova et al., CARF and WYL domains: ligand-binding regulators of prokaryotic defense
systems. Front Genet, 2014. 5: p. 102.
- 9. Cao et al., Identification and functional study of type III-A CRISPR-Cas systems in
clinical isolates of Staphylococcus aureus. Int J Med Microbiol, 2016.
- 10. Elmore et al., Bipartite recognition of target RNAs activates DNA cleavage by the
Type III-B CRISPR-Cas system. Genes Dev, 2016. 30(4): p. 447-59.
- 11. Estrella et al. RNA-activated DNA cleavage by the Type III-B CRISPR-Cas effector complex.
Genes Dev, 2016. 30(4): p. 460-70.
- 12. Samai et al., Co-transcriptional DNA and RNA Cleavage during Type III CRISPR-Cas Immunity.
Cell, 2015. 161(5): p. 1164-74.
- 13. Goldberg et al., Conditional tolerance of temperate phages via transcription-dependent
CRISPR-Cas targeting. Nature, 2014. 514(7524): p. 633-7.
- 14. Liu et al., RNA and DNA Targeting by a Reconstituted Thermus thermophilus Type III-A
CRISPR-Cas System. PLoS One, 2017. 12(1): p. e0170552.
- 15. Han et al., A type III-B CRISPR-Cas effector complex mediating massive target DNA
destruction. Nucleic Acids Res, 2016.
- 16. Tamulaitis et al. Type III CRISPR-Cas Immunity: Major Differences Brushed Aside. Trends
Microbiol, 2017. 25(1): p. 49-61.
- 17. Zhu and Ye, Crystal structure of Cmr2 suggests a nucleotide cyclase-related enzyme
in type III CRISPR-Cas systems. FEBS Lett, 2012. 586(6): p. 939-45.
- 18. Cocozaki et al., Structure of the Cmr2 subunit of the CRISPR-Cas RNA silencing complex.
Structure, 2012. 20(3): p. 545-53.
- 19. Makarova et al., A DNA repair system specific for thermophilic Archaea and bacteria
predicted by genomic context analysis. Nucleic Acids Res, 2002. 30(2): p. 482-96.
- 20. Hatoum-Aslan et al., Genetic characterization of antiplasmid immunity through a type
III-A CRISPR-Cas system. J Bacteriol, 2014. 196(2): p. 310-7.
- 21. Osawa, T., H. Inanaga, and T. Numata, Crystal structure of the Cmr2-Cmr3 subcomplex
in the CRISPR-Cas RNA silencing effector complex. J Mol Biol, 2013. 425(20): p. 3811-23.
- 22. Opoku-Temeng, C., et al., Cyclic dinucleotide (c-di-GMP, c-di-AMP, and cGAMP) signalings
have come of age to be inhibited by small molecules. Chem Commun (Camb), 2016. 52(60):
p. 9327-42.
- 23. Koonin, E.V. and K.S. Makarova, CRISPR-Cas: evolution of an RNA-based adaptive immunity
system in prokaryotes. RNA Biol, 2013. 10(5): p. 679-86.
- 24. Makarova, K.S., et al., Unification of Cas protein families and a simple scenario
for the origin and evolution of CRISPR-Cas systems. Biol Direct, 2011. 6: p. 38.
- 25. Makarova, K.S., et al., A putative RNA-interference-based immune system in prokaryotes:
computational analysis of the predicted enzymatic machinery, functional analogies
with eukaryotic RNAi, and hypothetical mechanisms of action. Biol Direct, 2006. 1:
p. 7.
- 26. Makarova, K.S., et al., Evolution and classification of the CRISPR-Cas systems. Nat
Rev Microbiol, 2011. 9(6): p. 467-77.
- 27. Wiedenheft, B., S.H. Sternberg, and J.A. Doudna, RNA-guided genetic silencing systems
in bacteria and archaea. Nature, 2012. 482(7385): p. 331-8.
- 28. Kim, Y.K., Y.G. Kim, and B.H. Oh, Crystal structure and nucleic acid-binding activity
of the CRISPR-associated protein Csx1 of Pyrococcus furiosus. Proteins, 2013. 81(2):
p. 261-70.
- 29. Niewoehner, O. and M. Jinek, Structural basis for the endoribonuclease activity of
the type III-A CRISPR-associated protein Csm6. RNA, 2016. 22(3): p. 318-29.
- 30. Anantharaman, V., et al., Comprehensive analysis of the HEPN superfamily: identification
of novel roles in intra-genomic conflicts, defense, pathogenesis and RNA processing.
Biol Direct, 2013. 8: p. 15.
- 31. Jiang, W., P. Samai, and L.A. Marraffini, Degradation of Phage Transcripts by CRISPR-Associated
RNases Enables Type III CRISPR-Cas Immunity. Cell, 2016. 164(4): p. 710-21.
- 32. Xiao, T.S. and K.A. Fitzgerald, The cGAS-STING pathway for DNA sensing. Mol Cell,
2013. 51(2): p. 135-9.
- 33. Danilchanka, O. and J.J. Mekalanos, Cyclic dinucleotides and the innate immune response.
Cell, 2013. 154(5): p. 962-70.
- 34. Corrigan, R.M. and A. Grundling, Cyclic di-AMP: another second messenger enters the
fray. Nat Rev Microbiol, 2013. 11(8): p. 513-24.
- 35. Silverman, R.H., A scientific journey through the 2-5A/RNase L system. Cytokine Growth
Factor Rev, 2007. 18(5-6): p. 381-8.
- 36. Gootenberg, J.S., et al., Nucleic acid detection with CRISPR-Cas13a/C2c2. Science, 2017.
- 37. Zheng, L., U. Baumann, and J.L. Reymond, An efficient one-step site-directed and site-saturation
mutagenesis protocol. Nucleic Acids Res, 2004. 32(14): p. e115.
- 38. Zhang, B.Z., et al., An easy-to-use site-directed mutagenesis method with a designed
restriction site for convenient and reliable mutant screening. J Zhejiang Univ Sci
B, 2009. 10(6): p. 479-82.
- 39. Yoshioka, K., KyPlot - A User-oriented Tool for Statistical Data Analysis and Visualization.
CompStat, 2002. 17(3): p. 425-437.
- 40. Sheppard et al. The CRISPR-associated Csx1 protein of Pyrococcus furiosus is an adenosine-specific
endoribonuclease, RNA, 22 (2016) p. 216-224
- 41. Lintner, N.G, Frankel, K.A, Tsutakawa, S.E, Alsbury, D.L, Copié, V, Young, M.J, Tainer,
J.A, Lawrence, C.M., The structure of the CRISPR-associated protein Csa3 provides
insight into the regulation of the CRISPR/Cas system, 2011.
- 42. Engl C, Schaefer J, Kotta-Loizou I, Buck M., Cellular and molecular phenotypes depending
upon the RNA repair system RtcAB of Escherichia coli, 2016
- 43. Tanaka N, Meineke B, Shuman S., RtcB, a novel RNA ligase, can catalyze tRNA splicing
and HAC1 mRNA splicing in vivo, 2011
- 44. Chakravarty A.K., Subbotin R., Chait B.T. and Shuman, S., RNA ligase RtcB splices
3'-phosphate and 5'-OH ends via covalent RtcB-(histidinyl)-GMP and polynucleotide-(3')pp(5')G
intermediates, 2012
- 45. Genschik P., Drabikowski,K.,and Filipowicz, W., Characterization of the Escherichia
coli RNA 3'-terminal phosphate cyclase and its sigma 54-regulated operon, 1998.
[0091] The embodiments shown and described in the specification are only specific embodiments
of inventors who are skilled in the art and are not limiting in any way.